Assessing the C/O Ratio Formation Diagnostic: A Potential Trend with Companion Mass

The carbon-to-oxygen (C/O) ratio in an exoplanet atmosphere has been suggested as a potential diagnostic of planet formation. Now that a number of exoplanets have measured C/O ratios, it is possible to examine this diagnostic at a population level. Here, we present an analysis of currently measured C/O ratios of directly imaged and transit/eclipse planets. First, we derive atmospheric parameters for the substellar companion HD 284149 AB b using data taken with the OSIRIS integral field spectrograph at the W.M. Keck Observatory and report two non-detections from our ongoing imaging spectroscopy survey with Keck/OSIRIS. We find an effective temperature of $T_\mathrm{eff} = 2502$ K, with a range of 2291-2624 K, $\log g=4.52$, with a range of 4.38-4.91, and [M/H] = 0.37, with a range of 0.10-0.55. We derive a C/O of 0.59$^{+0.15}_{-0.30}$ for HD 284149 AB b. We add this measurement to the list of C/O ratios for directly imaged planets and compare them with those from a sample of transit/eclipse planets. We also derive the first dynamical mass estimate for HD 284149 AB b, finding a mass of $\sim$28 $M_\mathrm{Jup}$. There is a trend in C/O ratio with companion mass ($M_{\mathrm{Jup}}$), with a break seen around 4$M_{\mathrm{Jup}}$. We run a Kolmogorov-Smirnov and an Anderson-Darling test on planets above and below this mass boundary, and find that they are two distinct populations. This could be additional evidence of two distinct populations possibly having two different formation pathways, with companion mass as an indicator of most likely formation scenario.